An Autofocus Method for Backprojection Algorithm in Range Compression Domain

Backprojection (BP) algorithm is the classic time-domain imaging method for synthetic aperture radar (SAR). The quality of BP imaging results relies on the accuracy of motion measurement, for which autofocus is often important in BP algorithm for airborne imaging. Traditional autofocus methods for frequency domain SAR imaging algorithms, such as phase gradient autofocus (PGA) can hardly be applied in BP algorithm. The most studied autofocus methods for BP, such as the maximum sharpness method or generalized phase gradient autofocus method, need large amount of memory or computation. In this work, we present an autofocus algorithm which is achieved before imaging for BP algorithm. The approach is based on the feature of strong scatterer in range compression domain, from which we can estimate the real trajectory of antenna phase centers and compensate the corresponding motion error and Doppler modulation phase. This algorithm holds the same concept as PGA, but it is applied before imaging to utilize the advantage of BP algorithm. The algorithm can be applied for wide scene with little computational amount. Simulation and experimental results are demonstrated to prove the efficacy of the algorithm.